Skip to content

Preprints

New results on bioRxiv and medRxiv, before peer review.

100 selectedRelated topicsOriginal researchMethods and resourcesSingle-cell and spatial omics

Latest picks

21–40 of 100
Oct 11Sun
  1. bioRxiv Genomics62

    Gene-environment interaction loci with signatures of mid- to late Holocene selection show evidence for diet-dependent genetic effects

    Three of 45 independent gene-environment interaction loci show significant signatures of directional selection, linked to urate, LDL-C and cholesterol levels, with parallel frequency increases and acceleration points around 4,700-7,200 years ago. Selection-associated GxE loci were enriched for urate-related variants and interactions involving meat and cheese consumption, with shared frequency increases across worldwide populations. The findings indicate gene-environment interactions are a previously underappreciated component of recent human evolution.

    Why it matters: It links widespread gene-environment interactions to human selection over the last ~10,000 years, reframing diet as an evolutionary force.

  2. bioRxiv Genomics62

    Spatiotemporal AI modeling of gene regulation and variant effects in the developing human brain

    Seal, an interpretable AI transfer learning framework, models gene expression and variant effects with spatiotemporal resolution across 26 brain regions, 30 cell types, and seven developmental stages of the human brain. Applied to GWAS, it identifies cell types and developmental windows relevant to six neuropsychiatric conditions, and in Simons Simplex Collection autism whole genome data it uncovers a significant burden of de novo regulatory variants in transient fetal excitatory neurons.

    Why it matters: It links disease-associated variants to specific cell types and developmental windows in the developing brain.

  3. bioRxiv Genomics61

    Cellular and molecular census of the pleiotropic disease locus Negr1 in zebrafish

    A knock-in reporter at the zebrafish Negr1 locus, combined with single-cell transcriptomics, whole-brain registration and single-neuron projection mapping, maps where this pleiotropic disease locus acts. Negr1 expression is graded across neuronal classes of brain and retina, radial glia, astrocyte-like cells and, most consistently, oligodendrocyte progenitor cells, with projection mapping suggesting three conserved hubs: olfactory bulb projection neurons, tectal periventricular projection neurons and cerebellar eurydendroid cells. The results support dose perturbation of a broadly expressed gene as the source of Negr1 pleiotropy across BMI, major depression, schizophrenia, autism-spectrum conditions and anorexia.

    Why it matters: It gives a cell-type atlas for a locus tied to BMI, depression, schizophrenia, autism and anorexia, informing which cells each trait may arise from.

  4. bioRxiv Genomics67

    Ayahuasca plant genomes reveal retrotransposon-derived centromeres and contrasting routes to DMT and β-carbolines

    Genome assemblies of the two ayahuasca source plants reveal contrasting histories: Psychotria viridis is an autopolyploid hexaploid (2n = 6x = 66) while Banisteriopsis caapi is diploid (2n = 2x = 20), yet centromeres of both are enriched in CRM-type Ty3/Gypsy retrotransposons. In P. viridis a single multifunctional methyltransferase catalyzes both N-methylations of tryptamine to DMT, whereas B. caapi makes beta-carbolines via dedicated enzymes plus a non-enzymatic, acetaldehyde-dependent Pictet-Spengler cyclization followed by enantioselective oxidation to harmaline. The work supplies the genetic structure of traditionally recognized cultivated varieties and the enzymes behind the alkaloids combined in the decoction.

    Why it matters: First genomes of both ayahuasca plants identify the enzymes making DMT and beta-carbolines and the basis of cultivated varieties.

  5. bioRxiv Genomics68

    Disrupted chromatin architecture blocks B lymphopoiesis in Down syndrome

    Three-dimensional genome architecture organizes human fetal haematopoiesis and explains the B-lineage defect in Down syndrome. Using ChromE-seq, a new single-cell assay jointly profiling chromatin conformation and transcription, the authors built a multimodal atlas of ~400,000 haematopoietic cells from disomic and trisomy 21 fetal liver. Trisomy 21 disproportionately disrupted B-lineage architecture, weakened an EBF1-centred regulatory circuit, and deleting a T21-sensitive EBF1 loop anchor reproduced the CLP-to-B-progenitor bottleneck seen in T21.

    Why it matters: Links 3D chromatin architecture to lineage priming and identifies an EBF1 circuit behind the trisomy 21 B-cell defect.

  6. bioRxiv Genomics71

    Massively scaled variant effect mapping across the MAPK pathway reveals the importance of molecular and cellular context

    Variant effect mapping of 186,676 variants across 17 RAS-MAPK pathway proteins in human cells, totaling 503,099 measurements of protein abundance and pathway activity, revealed 14,412 dominant negative variants across 12 proteins, nearly all previously unknown and depleted from the human population. HSP90 inhibition experiments showed HSP90 shapes the effect of 67% of missense variants across the pathway. The data also resolved how the KRAS switch II helix surface accommodates overlapping LZTR1 and SOS1 interactions through distinct allosteric networks.

    Why it matters: Because it shows cellular context, especially HSP90, shapes the effect of most variants in a core signaling pathway.

  7. bioRxiv Genomics70

    A unified framework for quantitative splicing and transcript prediction

    SpliceAI2 is a deep learning model trained end-to-end on splice site, splice junction, and transcript usage data across multiple mammalian species, and it reconstructs complete transcripts directly from genomic sequence. It improves detection and quantification of splice-altering variants across benchmarks covering cryptic splicing, splicing QTLs, and massively parallel reporter assays. In a rare disease cohort, predicted splice-altering variants account for 15% of the excess genetic burden, with about half in deep intronic regions likely missed by exome sequencing.

    Why it matters: SpliceAI2 quantifies splice-altering variant effects genome-wide, including deep intronic variants exome sequencing misses.

  8. bioRxiv Genomics69

    Scalable methylome-transcriptome profiling resolves clonal organization in human hematopoiesis

    Droplet-MT jointly profiles transcriptomes and methylomes at ~10,000 cells per library, using methylation epimutations for mutation-independent clonal reconstruction and label transfer at ~0.2% CpG coverage to cut sequencing cost. Applied to ~21,000 paired profiles from healthy blood and longitudinal acute myeloid leukemia, it resolved small and expanded clones with heterogeneous lineage output and distinguished malignant clonal abundance changes, state remodeling within persistent clones, and recovery of healthy polyclonal hematopoiesis. Malignant monocytes stayed transcriptionally and epigenetically distinct from healthy and recovery-associated cells despite treatment-associated remodeling.

    Why it matters: It offers a scalable way to link clonal ancestry to molecular state in human hematopoiesis and leukemia.

  9. bioRxiv Genomics67

    An iPSC-based experimental framework for investigating phenotypic variation associated with Jomon ancestry in the present-day Japanese population

    Jomon ancestry, measured by the Jomon allele score, tracked with renal-related traits and serum triglycerides in 34,152 Tohoku Medical Megabank participants, and donors at JOM extremes (n = 10) showed group-level differences in immune-related gene expression. But when hiPSCs from those donors were differentiated into hepatocyte-like cells, intracellular triglycerides and lipid-related gene expression did not differ between groups. The work establishes a genetically characterized hiPSC framework for testing ancestry-associated phenotypes at the cellular level.

    Why it matters: It links population-scale ancestry associations to direct cellular tests using donor hiPSC-derived hepatocyte-like cells.

Oct 10Sat
  1. arxiv.org40

    Learning Probabilistic Logic Programs with Functional Gradient Guided Language Models

    grasp (Gradient-boosted Synthesis of Probabilistic logic programs) learns probabilistic logic programs by treating relational structure learning as functional gradient boosting, with first-order rules as weak learners and an LLM serving as the proposal oracle for the inner search. Across four relational benchmarks, including molecular toxicity prediction (Tox21), mutagenesis, and citation matching (Cora), grasp improves over purely symbolic, neural, and LLM-based baselines while yielding interpretable weighted rule ensembles and retaining boosting guarantees.

    Why it matters: It shows LLM-guided boosting can synthesize interpretable logic programs that beat symbolic, neural, and LLM baselines on molecular benchmarks.

  2. bioRxiv Genomics62

    Systematic mapping of enhancers, silencers, and topological elements with prime editing deletion screens

    Swap-seq, a method combining pooled twin prime editing with flow-sorting, deletes hundreds of endogenous regulatory sequences and measures their quantitative effects on nearby gene expression. Applied to PPIF regulation in THP-1 monocytes, it quantifies distal and intronic enhancers, characterizes CTCF insulators, and uncovers a silencer that represses neighboring enhancers through HDAC-mediated mechanisms, with hundreds of genomic elements sharing its signatures. The resulting data show sequence-to-function models capture local element activity but fail to link elements to distal target genes.

    Why it matters: It delivers a generalizable deletion-screening tool and a benchmark showing sequence-to-function models miss distal gene targeting.

Oct 9Fri
  1. bioRxiv Genomics66

    Random mtDNA mutagenesis maps genotype-phenotype associations across metabolic contexts and lineages

    mtDIVE, a random C→T mutagenesis platform, enables pooled mitochondrial forward genetics in human epithelial cells and zebrafish embryos, generating molecules from single substitutions to more than 100 mutations. Selection on mutant genomes and their transcriptional consequences depend on metabolic state and developmental lineage: obligate OXPHOS eliminates highly mutated cells by apoptosis, while glycolytic conditions promote their competitive displacement. Individual loci have modest, non-redundant effects that predict cell state beyond mutation burden, and cancer-recurrent variants are enriched among mutations linked to reduced mitochondrial transcription.

    Why it matters: It replaces one-allele-at-a-time mtDNA models with pooled forward genetics separating mutation identity from load.

  2. bioRxiv Cancer Biology65

    The Red Blood Cell Surface Defines a Distinct Tumor DNA Compartment for Liquid Biopsy

    Red blood cells tether and retain tumor-derived DNA on their surface, forming a distinct circulating pool of RBC-bound tumor DNA (rtDNA), with desialylation increasing capture in a Toll-like receptor 9-dependent manner. In mice bearing lung cancer xenografts, rtDNA was detected in all animals and tracked tumor burden while plasma tumor DNA was undetectable; in patients with lung cancer, most detected mutant tumor DNA was RBC-associated and concordant with matched tumor tissue. The findings challenge the plasma-centric liquid biopsy paradigm by identifying the RBC surface as a tunable reservoir of tumor DNA.

    Why it matters: It identifies red blood cells, not plasma, as the main circulating compartment of tumor DNA for liquid biopsy.

  3. bioRxiv Genomics60

    STEER-seq: a recombinase based platform for pooled multimodal single-cell perturbation screening

    STEER-seq is a virus-free platform for pooled ORF gain-of-function screening in human induced pluripotent stem cells, using Bxb1-mediated integration at a defined safe-harbour locus so each perturbation is expressed from a single copy in a common genomic context. Joint RNA and chromatin accessibility profiling linked transcription factor responses to chromatin remodelling. In a haematopoietic screen, SPI1 alone drove differentiation to HSC/MPP-like progenitors without exogenous cytokines, and scaling to around 2,000 human coding sequences uncovered factors inducing neural progenitor-, melanocyte- and vascular smooth muscle-like states.

    Why it matters: It removes random viral integration as a confound in pooled gain-of-function screens and works in hiPSCs at ~2,000-ORF scale.

Oct 2Fri
  1. bioRxiv Genomics67

    Genome-wide mapping of endogenous topoisomerase cleavage complexes reveals TOP1 at the interface of DNA topological regulation and genome instability

    Protein-DNA adduct immunoprecipitation followed by sequencing (pDAIP-seq) detects endogenous topoisomerase cleavage complexes genome-wide without stabilization by topoisomerase poisons, revealing that camptothecin, required by prior TOP1cc profiling methods, substantially alters the genomic distribution of TOP1cc. Endogenous TOP1cc were reduced by transcriptional suppression and increased by TOP2A depletion, and their positioning was largely specified by local DNA sequence context. The method also maps TOP2cc, TOP3cc and SPO11-DNA intermediates in mouse testis and identifies endogenous TOP1cc as a potential major source of spontaneous DNA breaks in human cells.

    Why it matters: pDAIP-seq maps endogenous topoisomerase cleavage complexes genome-wide without poisons, changing how TOP1cc distributions are measured.

  2. bioRxiv Cancer Biology59

    KRAS inhibition drives adaptive surfaceome remodeling and reveals ALPP/ALPPL2 as a targetable vulnerability in pancreatic cancer

    Quantitative temporal profiling of the global, phosphorylation, and surface proteome in PDAC shows that KRAS inhibition upregulates the cell surface protein ALPP through compensatory JAK2/STAT3 signaling. The authors developed a clinical-grade ALPP/ALPPL2-targeting antibody drug conjugate with site-specific MMAE conjugation, which showed single-agent efficacy in PDAC models and enhanced internalization and anti-tumor efficacy combined with KRAS inhibitors, pointing to a combination strategy to forestall KRASi resistance.

    Why it matters: It defines the KRASi-induced proteome and links a surface target to a clinical-grade ADC combination strategy against resistance.

  3. bioRxiv Genomics53

    Phenotypes of ultra-rare variant carriers benchmark variant effect scores

    UKBBGym benchmarks coding and non-coding variant effect scores against plasma protein abundance and quantitative traits in ultra-rare variant carriers in the UK Biobank. Scores designed for coding and splicing variants outperform transcriptional-regulation scores on these phenotypes, and experimental assays do not consistently beat computational scores for missense effects. The benchmark is reproducible from public summary statistics for coding variants and avoids ascertainment bias of clinical labels.

    Why it matters: It supplies a population-phenotype benchmark for variant effect scores that avoids clinical-label ascertainment bias.

  4. bioRxiv Genomics70

    The Genome In A Bottle HG002 assembly-based variant benchmark set enables comprehensive benchmarking of small and structural variants.

    The Genome in a Bottle Consortium's HG002 v5.0q benchmark set, built from a fully phased, highly curated diploid assembly, delivers confident small- and structural-variant calls in previously inaccessible regions such as segmental duplications, tandem repeats, homopolymers, and complex immune loci. Relative to v4.2.1 it adds 200 Mbp of callable sequence, increases variant detection by 18% for small variants and 3-fold for structural variants, and is available for GRCh37, GRCh38, and CHM13.

    Why it matters: It sets a new standard resource for validating sequencing technologies and variant callers in regions older benchmarks miss.

Oct 1Thu
  1. bioRxiv Cancer Biology66

    Transcriptional rewiring of NF-κB by HBZ and Fra-2 drives inflammatory chemoresistance in Tax-negative adult T-cell leukemia/lymphoma

    HTLV-1 antisense protein HBZ stabilizes a Fra-2/p65 complex that rewires NF-κB transcription toward IL-1β in tax-negative adult T-cell leukemia/lymphoma, driving inflammatory chemoresistance. IL-1β was elevated in ATL patient sera and cells independently of proviral load, and blocking IL-1 signaling with anakinra or anti-IL-1β antibodies restored doxorubicin sensitivity in primary ATL cells. The work identifies the Fra-2-NF-κB-IL-1β axis as a targetable driver of chemoresistance.

    Why it matters: It identifies a druggable IL-1 signaling axis behind chemoresistance in tax-negative ATL.

  2. bioRxiv Cancer Biology64

    Early dynamics of a neutrophil interferon program drive immunotherapy resistance and limit durable anti-tumor immunity

    An interferon-stimulated gene program in neutrophils, detectable before tumors diverge in their response, predicts resistance to anti-PD1 therapy in preclinical triple-negative breast cancer models. Persistent neutrophil IFN signaling in blood and tumor tracked with treatment failure, while genetic disruption of type I or type II interferon signaling in neutrophils improved anti-PD1 efficacy and durable anti-tumor immune memory. The work positions IFN-responsive neutrophils as blood-based biomarkers and therapeutic targets.

    Why it matters: Identifies an early, blood-accessible neutrophil IFN signature that predicts anti-PD1 resistance and can be targeted.

CBACBA Frontier Insights (CFI)Chinese Biopharmaceutical Association (CBA)www.cba-usa.org